This community innovation has been accepted at the 2025 DHIS2 Annual Conference
PRIDEC: Forecasting diseases in DHIS2 and R
Climate change is impacting both long-term climate trends and short-term extreme climatic events, making infectious disease dynamics increasingly difficult to predict. The wide availability of remotely-sensed environmental data and public health metrics has resulted in the proliferation of climate-sensitive disease forecasting tools, which can aid health systems in adapting to climate change, but few are available for health actors to use themselves. The Madagascar Ministry of Health and Pivot, a non-governmental organization working in health system strengthening, have collaborated to create a disease forecasting tool and associated suite of open-source software (Predicting Infectious Diseases via Environment and Climate; PRIDE-C), including a DHIS2 application and complementary R package. The PRIDEC R package allows the user to create disease forecasting models from DHIS2 health and climate data. It includes modules to tune, fit, and forecast from multiple statistical modeling frameworks and outputs interactive reports to interpret and validate the models. In this presentation, we will report on the development of these tools through a series of participatory workshops and present the results of applying this package to forecast three diseases (malaria, diarrheal disease, and respiratory infections) in Ifanadiana district, southeastern Madagascar. We will discuss the predictive abilities and limitations of these models when applied to multiple diseases, as well as inferences regarding climate-health associations for each disease in this region.
Primary Author: Michelle Evans
Keywords:
R programming language; Epidemiological forecasting; climate; infectious disease